xtreg with fe option

[Yun Cheng]

Dear statalist,

I want to run a regression with firm and year fixed effects, and with standard errors clustered at the firm level. I did the following:

xtreg y x controls i.year i.firm, fe

The regression shows that

Fixed-effects (within) regression
Group variable: firm

I'm not familiar with xtreg, does Group variable: firm indicate that standard errors are clustered at the firm level? Thanks!

[Clyde Schechter]

No. Group variable: firm indicates that the firm variable is used as the fixed effects. (The inclusion of i.firm in the command is redundant--that is automatically taken care of by the -fe- after -xtset-, and I think you will see in your Results that the i.firm variables are all omitted due to colinearity (with themselves!). If you want clustered standard errors, you need to specify that in the -vce()- option.

xtset firm
xtreg y x controls i.year, fe vce(cluster firm)

[Yun Cheng]

I see, thanks Clyde! Yes, if I wrote in the way shown in post no.1 then all firm variables are omitted.